No so called experts worldwide are covering their butts by over stating the numbers. We were told 1 million deaths in the USA a month ago, and that “expert” has now down graded that number to 100,000. So forgive me if I take these speculations with a grain of salt.
You aren't paying attention. It was way more than that at 1.7 million, not 1, and that was a model based on NO 'gubmint' intervention, as some idiots here at FR seem to advocate. Do you think "worst case model'" distinction might be important? What part of "worst case model" didn't you understand?
Those were AI models based on past epidemics, very complex data mining, and considering the unique characteristics of the Kung Flu, they were believable.
Fortunately our president and his team are mitigating those unmitigated numbers, in a bigly way.